UH

AgueraArakilArtibaiBaiaBarbadunBidasoaButroeDebaEbroEgaIbaizabalInglaresJereaKarrantzaLeaOiartzunOkaOmecil loOriaPurónUrolaUrumeaZadorra

Number of claims (#)

2,000 4,000 8,000 12,000

Euros (€)

0 mln to 20 mln20 mln to 40 mln40 mln to 60 mln60 mln to 80 mln80 mln to 100 mln100 mln to 120 mln120 mln to 140 mln

S a n t i a g o G a z t e l u m e n d i 1 , 2 , J o s e b a E g a ñ a 1 , 2 , J o s é A n t o n i o A r a n d a 1 , 3

1- Basque Meteorology Agency (EUSKALMET). Parque tecnológico de Álava. Avda. Einstein 44 Ed. 6 Of. 303, 01510 Miñano, Álava, Spain2- TECNALIA, Meteorology Area. Parque tecnológico de Álava. Avda. Albert Einstein 28, 01510 Miñano, Álava, Spain.

3- Emergencies and Meteorology Directorate, Security Department, Basque Government, Basque Country, Spain.

Contact info:

santiago.gaztelumendi@tecnalia.com

TECNALIA

Parque Tecnológico de BizkaiaC/ Geldo Edificio 700E-48160 DERIO (Bizkaia) Spainwww.tecnalia.com

An analysis of f looding economic impact in Urumea area

AcknowledgementsThe authors would like to thank the Emergencies and Meteorology Directorate - Security Department - BasqueGovernment for public provision of data and operational service financial support and Spanish Insurance CompensationConsortium for data provision. We also would like to thank all our colleagues from EUSKALMET and URA for their dailyeffort in promoting valuable services for the Basque community. Finally, we would also like to thank the Free SoftwareCommunity and all institutions that maintain and support availability of free data and tools for the Scientific Community.

Severe weather phenomena impact the Basque society and economy in many ways, from disruption in various sectors and substantial damages in infrastructure to human and economic losses. Particularly flooding is the natural event that causes

the most disaster damage in Basque Country.

In this work we focus on flood impact produced in Urumea area during this century. Urumea river is sited in the East part of Basque Cantabric Basin, flowing to the Bay of Biscay through the city of San Sebastian. Is a representative river for this part

of the country, with rather steep slope and short concentration time. In Urumea basin high precipitation events, flash-floods and associated impact are relatively usual.

In this study we use, as a reference for impact, the losses from flood damage paid by Spanish Insurance Compensation Consortium (CCS) during last twenty years. Data are analyzed and presented in different ways grouped per date, period, amount, damage type, location, etc. The final objective is to contribute to knowledge of impact characteristics in this area,

increasing awareness and preparedness in flash-floods events.

..

Abstract

IntroductionIn the Basque Country, north of the iberian Peninsula, rivers flows in two main watersheds, the

Cantabric and the Mediterranean (see Fig. 1 and 2). In the Cantabric basins high precipitations

rates and steep slopes of around 1000 meters in few kilometers promotes flooding with certain

regularity. Nevertheless, the rivers of the Mediterranean slope have less incidence rate of

floods due to less precipitation rates and slopes (see Fig 2). Is important to note that

Mediterranean basins are mainly in the territory of Alava the largest (2,963 km²) of the three

provinces in the Basque Autonomous Community but the less industrial, inhabited and flat part

of the territory. On the other hand in Bizkaia (2.217 km², 1.141.000 inhabitants) and Gipuzkoa

(1980 km²), 700.000 inhabitants ) are occupied by many river´s valleys highly populated and

with industrial activities. If we consider those different geographical and sociological factors, it

is not strange that 95% of the economic impact of floods occurs in the Cantabrian basin and

particularly more than 30% in Ibaizabal basin (1.800 km² ) and an amount slightly lower in

Urumea Basin with six times less extension (270 km²) (see Fig 2).

The Urumea river is 56,91 km long from its source in a mountainous area of Ezkurra (695 m) to

the city of Donostia-SanSebastian (capital of Gipuzkoa), its catchment has 272,4 km². Half the

length of the river and one third of the total area of its basin are outside Gipuzkoa, in Nafarroa

Urumea basin has two distinct parts: the highest part of the basin, with very little human

occupation and the heavily modified and altered zone that begins in Hernani and extends to

Astigarraga and its mouth in San Sebastian. The average slope of the main course is 1,26%,

reaching 16,45% in higher areas. In some sections both the main course and minor

components, especially in the upper basin has a high torrentiality (see Fig 1). The lower stretch

of the river has wide and flat banks fit for crops and urban development, which has resulted in

the most inhabited drainage basin in Gipuzkoa, (213.000 inhabitants). The main towns on the

river are Hernani (19.700 inhabitants), with the first industrial estates upstream locating on its

banks, Astigarraga (3.800), renowned for its cider houses, and a final long stretch snaking

through different parts of Donostia-SanSebastian (186.500 inhabitants) (see Fig 1).

Is important to note that CCS gives compensation for the damages produced by Extraordinary

Risks (natural phenomena and events of a political or social nature), on the condition of holding

a policy in the field of damages to goods, or life and/or accidents, with any insurance company..

When giving compensation, the CCS will take into account the same goods, the same insured

capital and any other conditions (first loss, compensation limits, etc.) established in this

insurance policy for the aforementioned contingencies.

Remarks & conclusions

• CCS, web del consorcio de compensacion de seguros. http://www.consorseguros.es/web/inicio.

• DAEM. 2015. Caracterización y estudio meteorológicode situaciones de emergencia porprecipitaciones (Urumea y Kadagua). Internal final documentation project. Coord. S. Gaztelumendi. Tecnalia contract no: 050597.

• EUSKALMET 2016. Basque Meteorology Agency web. www.euskalmet.euskadi.eus• EUSTAT 2016. Basque Statistical Office (EUSTAT) web , www.eustat.eus/• Egaña, J., Gaztelumendi, S. Palacio, V., Martija, M. 2015. A case study of local floods in

Urumea area (Basque Country). 15th EMS Annual Meeting 12th European Conference on Applications of Meteorology (ECAM) 07 – 11 September 2015, Sofia, Bulgaria.

• Egaña, J., Gaztelumendi, S., Gelpi, I.R., Mugerza, I. 2005. Synoptic patterns associated to very heavy precipitation events in the Basque Country. EMS5/ECAM7 Utrech Netherland.

• Egaña, J., Gaztelumendi, S., Gelpi, I.R., Otxoa de Alda, K., Maruri, M., Hernández, R., 2008: Radar Analysis of Different Meteorological Situations in the Basque Country Area. ERAD 2008, Helsinki, Finland.

• Egaña, J., Gaztelumendi, S., Pierna, D., Gelpi, I.R., Hernández, R., Otxoa de Alda, K., 2010: Flash-floods on Basque Country at the end of January 2009. EMS10/ECAC8. Zurich, Switzerland.

• Egaña, J., Gaztelumendi, S., Palacio, V., Gelpi, I.R., Otxoa de Alda, K., 2012: Using Euskalmet Radar data for analysis of a persistent precipitation case. ERAD 2012. Toulouse, France.

• Egaña, J., Gaztelumendi, S., Ruíz, M., Hernández, R., Gelpi, I.R., Otxoa de Alda, K., 2013: A study of a flood episode in Basque Country. ECSS 2013, Helsinki, Finland.

• Egaña J, Gaztelumendi S. Aranda, J.A.: 2016. Meteorological characterization of events that generate floods impact in north part of Basque Country. 16th EMS /11th ECAC 12–16

September 2016 , Trieste, Italy• Gaztelumendi S. Egaña J, Aranda, J.A.: 2016. Analysis of coastal impact in Basque

Country . 16th EMS /11th ECAC 12–16 September 2016 , Trieste, Italy• Gaztelumendi, S., Egaña, J., Pierna, D., Palacio, V., Gelpi, I.R., Hernández, R., Otxoa de

Alda, K., 2011: A study of a flashflood case in Basque Country: the 16th june 2010 event. EMS11/ECAM10. Berlin, Germany.

• Ibisate, A., Ollero, A., Ormaetxea, O., 2000: Las inundaciones en la vertiente cantábricadel País Vasco en los últimos veinte años: principales eventos, consecuencias territorialesy sistemas de prevención. Serie Geográfica, 9, Departamento de Geografía, Prehistoria y Arqueologia. Universidad del País Vasco.

• Lozano Valencia, P., 2006: Régimen precipitacionalen el norte de Navarra y Guipúzcoa¿Récord peninsular y europeo?. Nimbus, nº17-18, ISSN 1139-7136, 125-144

• Mugerza I, Gaztelumendi S, Egaña J, Gelpi IR: “Rainfall and snowmelt: a dangerous meld for flooding”. 5th Annual Meeting of the European Meteorological Society (EMS). Utrech(Netherlands), 12-16 September 2005

• Pejenaute Goñi, J., 2008: Inundaciones históricas en los valles cantábricos navarros(1881-2007). http://www.opcc-ctp.org/etudes/Inundaciones_hist__ricas_....pdf

• Prieto, C., Lamas, J.L., 1985: Avenidas extraordinarias en el País Vasco. Geología y prevención de daños por inundaciones. IGME, pp. 247-334.

• Ruíz, M., Egaña, J., Gaztelumendi, S., Maruri, M., Gelpi, I.R., 2012: A case study of heavy and persistent rainfall. ERAD 2012, Toulouse, France.

• R Core Team (2014). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.

• Tennekes, M., 2016. Package ‘tmap’. CRAN Repository 2016.• URA 2016. Basque Water Agency web. www.uragentzia.euskadi.eus/

References

DiscussionThe data analyzed correspond to the compensation of authorized claims for flood damages during the period 1996-2015 by the CCS. An amount

of 99.760.000 euros (€) corresponding to 9.441 accepted claims (#) during 108 different days (D) with some degree of damage impact.

Damages are negligible in the Nafarroa part of the Basin, and are produced mainly in three different Gipuzkoan municipalities. Donostia,

Hernani and Astigarraga registering 77% , 13% and 9% of #, corresponding to 72% , 9% and 19% of the economic amount (€) during

78, 35 and 43 days respectively. This is explained considering the particular characteristics of occupation, urban and industrial use of these

areas. Remark the highest ratio of 23.000 €/# in Hernani as a consequence of relative influence of industrial damages in this area.

If we consider the type of damage, we found that business (shops, stores, warehouses, etc) account for 18% of # with a value of 35% of €

produced during 52% of the days (56days). Industrial category account for 5% of #, 25% of € and 27% days (29), civil works 0.1% #, 2% € ,

10% days (11), offices 2% #, 16% € for 14% days (15), motor vehicles 20% #, 6% €, 26% days (28), housing (homes and communities) 56%

#,16% €,62 days (66). On average each # represents 10K€, 21K€ for business, 56K€ for industrial, 127€ for civil works, 91K€ for offices, 3K€

for vehicles and 3K€ for housing..

In business related claims (#) highlight a particular case for Donostiawhere more than 4.000K€ in damages where produced in a single

reclamation during 01/06/1997 in a warehouse. In Industrial claims stand out Donostia during 01/06/1997 with two # for more than 4.000K€. In civil works case damages are registered just in Donostia mainly due to railways interventions and particularly for one episode in 01/06/1997

(1.000K€). Regarding with offices , two claims during same day account for more than 12.000 K€ also in Donostia (82%). In motor vehicles and housing remark again Donostia with 87% and 82% of # corresponds to

83% and 78% of € respectively.

Although during last twenty years damages are produced during 108 different days, just during 21 days damages are register for more than

100.000 €, and during 20 days with more than 10 #. Two particular days account for 72% of # and 83% of total €. During 01/06/1997 and 06/11/2011 the 45% and 27% of # and 53% and 30% of € were

produced respectively.

In relation to the type of damage and date when is produced, regarding business 37%, 17% and 13% of # and 47%, 28% and 4% of € are

produced during 01/06/1997, 06/11/2011 and 25/0872002 days, respectively. In industrial case 47% and 30% of # and 44% and 45% of

€ during 01/06/1997 and 06/11/2011. For civil works all claims are produced in 11 days, highlighting 01/06/1997 day with 68%. In offices

case 01/06/1997 account for 58% of # and 94% of €. In vehiclesclaims, 58% in 01/06/1997 and 30% 06/11/2011 . In housing

01/06/1997 stand out for 42% # and 36% € and 06/11/2011 for 30% # and 44% €.

In fig 8. we can see the monthly and yearly distributions of #,€ and D. Remark the effects od two most harmful episodes. In both case with a

cut-off low in high levels. June episode driven by a MCS and intensity of precipitations, November case by a very active frontal event and

persistent precipitations(see details on meteorological characterization in Egaña et al 2016 – POSTER H62: EMS2016-311).

0

1020

30

40

Ag

üe

ra

Art

iba

i

Ba

rba

du

n

Bid

aso

a

Bu

tro

e

De

ba

Iba

iza

ba

l

Ka

rra

ntz

a

Lea

Oia

rtzu

n

Oka

Ori

a

Uro

la

Uru

me

a

Ara

kil

Ba

ia

Eb

ro

Eg

a

Ing

lare

s

Om

eci

llo

Za

do

rra

%

%#%€

7 7

0 0 0

45

28

0 1

28

23 40 0 0

54

14

0 1

33

1

1815

30 2

12

510

2

1016

7

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

0

45

0 0 0 0

9

1

6

0 0 1 0 1 0

28

0 1 1

6

0

54

0 0 0 0

5

03

0 0 1 0 1 0

33

0 1 03

6

10

4 37

1

97

2 1 15 4

7

0

7

0

12

7 7

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

# € D

Fig 8. Yearly and monthly distribution of #,€ and D in % for

period 1996-2015 (99.710.000 €, 9.400 accepted claims (#),

107 days (D))

Flood damages and economic impact result from the interplay of complex societal and hydro-meteorological factors.

In this work insurance claims data are analyzed for a selected Basin in the Basque Country, with a high concentration of

population and floods events. Urumea basin is representative of a small Cantabric river basin with high flood impact

potential, Urumea river flows through very populated and industrial areas.

During last 20 years we have108 different days (D) with accepted claims, but just 21 days damages are register for

more than 100.000 €, and during 20 days with more than 10 #. Two particular days account for 72% of # and 83% of total €.

The direct use of number of accepted claims (#) and economic quantity (€ )

seems to be useful as a proxy of impact and extension of damages,

particularly in the context of very severe events. But is not possible to use

directly as a measure of total losses. .Number of accepted claims (#) can be used to characterize the extension of damages produced in a given event,

better than €. Claims from Housing and Cars seems to be the better indication for extension of damages, especially

during less severe episodes.

ElevationLess than 00 to 5050 to 100100 to 150150 to 200200 to 250250 to 300300 to 350350 to 400400 to 450450 to 500500 to 750750 to 1,0001,000 to 1,2501,250 to 1,5001,500 or more

Municipality BUSINESS INDUSTRIAL CIVIL WORK OFICCES MOTOR VEHICLES HOUSING

ASTIGARRAGA # 211 81 13 128 800

K€ 4.280 2.258 - 88 338 1.964

DONOSTIA-SAN SEBASTIAN

# 1.283 228 14 154 1.528 4.098

K€ 22.734 12.623 1.785 15.499 5.175 13.480

EZKURRA # 1 1

K€ - - - - 3 1

GOIZUETA # 2

K€ - - - - - 6

HERNANI # 188 116 9 173 326

K€ 7.555 9.473 - 476 374 849

LEITZA # 2 1 1

K€ 40 2 - - 1 -

URNIETA # 21 16 1 14 30

K€ 505 159 - 5 35 51

ZUBIETA # 1

K€ - - - - - 0,5

1996-2015 # K€ D K€/# K€/D #/D

BUSINESS 1.705 35.114 57 21 616 30

INDUSTRIAL 442 24.516 29 55 845 15

CIVIL WORK 14 1.785 11 127 162 1

OFICCES 177 16.069 15 91 1.071 12

MOTOR VEHICLES 1.845 5.925 28 3 212 66

HOUSING 5.258 16.351 66 3 248 80

MEAN 1.574 16.627 34 11 484 34

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

# K€ D K€/# K€/D #/D

ZUBIETA

GOIZUETA

EZKURRA

LEITZA

URNIETA

HERNANI

ASTIGARRAGA

DONOSTIA-SANSEBASTIAN

0%

20%

40%

60%

80%

100%

# K€ D K€/# K€/D #/D

HOUSING

MOTOR VEHICLES

OFICCES

CIVIL WORK

INDUSTRIAL

BUSINESS

0%

20%

40%

60%

80%

100%

0%

50%

100%

URUMEA BASIN : 1996-2015 # K€ D K€/# K€/D #/D

GIPUZKOA 9.432 99.707 106 941 941 89

ASTIGARRAGA 1.233 8.928 43 208 208 29

DONOSTIA-SANSEBASTIAN 7.305 71.296 82 869 869 89

HERNANI 812 18.728 35 535 535 23

URNIETA 82 756 6 126 126 14

NAFARROA 9 52 6 9 9 2

EZKURRA 2 3 2 2 2 1

GOIZUETA 2 6 2 3 3 1

LEITZA 4 43 2 21 21 2

ZUBIETA 1 0 1 0 0 1

TOTAL/MEAN 9.441 99.760 108 924 924 87

URUMEA BASIN : 1996-2015

Affected days (D)Less than 11 to 55 to 1010 to 1515 to 2020 to 2525 to 5050 to 7575 to 100100 or more

Number of claims (#)

500 1,000 1,500 2,000 2,500

Euros (€)Less than 00 to 500,000500,000 to 1,000,0001,000,000 to 1,500,0001,500,000 to 2,000,0002,000,000 to 2,500,0002,500,000 to 3,000,0003,000,000 to 3,500,0003,500,000 to 4,000,0004,000,000 to 4,500,0004,500,000 to 5,000,0005,000,000 or more

#

5

20

50

100

200

500

1000

1500

10000

20000

30000

K€

MUNI

ASTIGARRAGA

DONOSTIA

EZKURRA

GOIZUETA

HERNANI

LEITZA

URNIETA

ZUBIETA

Fig 2. # and € for Basque

Basins during 1996-2015.

Fig 1. Geographic characteristics and municipalities in Urumea Basin.

Fig 3. Data summary focusing on municipalities.

Fig 4. Data summary focusing on type of property.

Fig 5. # and € summary for municipality and type.

Fig 6. D, # and € detailed by zip code.

Fig 7. # and € detailed by district (when

available) and municipality.

MethodologyOriginal data consist on a excel file from CCS with accepted claims

corresponding to “floods” for Basque Country during period 1996-2015. Information is structured in date, municipality, town, zip

code, risk type and economic amount. Original data present gaps (void cells), different type of inhomogeneity (different names for same place or municipality, language, etc.), inconsistencies (in between zip code, location and municipalities) and errors (e.g.

claims far away from littoral areas).

A derived data file is prepared after manual depuration of raw data solving most part of detected problems and extraction of claims for

Urumea Basin area based on location information (zip codes and municipality). Data analysis is performed applying different techniques after segmentation of data in various categories

considering type of affected property. Counting and statistics values are calculated focusing on #, €, D and M. Different R (R core Team

2013) and excel tools are used for depuration, data analysis, and reporting, including PivotTables and thematic maps.

Although the raw claim data contain location information (zip code, municipality and city name), often presents

inconsistencies . It would be useful to have georeferenced claims with exact location of damage. It would be very useful , especially when

particular spatial aggregations are needed, as

is the case for watershed/river damages

studies.

The CCS data represent just a portion of the total economic impact. Mostly due to the

different degree of penetration on insurance (density of the insurance and the economic

value of the property is not the same in every area) and to different considerations inherent

to the “extraordinary risk” compensation system (not all type of policy entitles to compensation and not all goods ,type of

damages or causes are covered). Usually different rates are applied depending on type of property insurance (e.g. Housing 0.09 per

thousand or certain civil works 1.95 per thousand).

.

In order to extract full conclusions about economic impact ,due to floods, using insurance claims data, many factors must be considered. Those affecting

potential exposure to flood event, as physical characteristics (river flow, river side conditions, etc.),

or socioeconomic aspects (distribution of human population and goods). Those affecting actions taken during episodes (early warning, dam operations, etc..).

Those related with preventive measures applied, previous damages or even chance. And those related

with the insurance politics and cultural aspects (characteristics and amount of the insured assets,

compensation policies, etc,). All of this factors must be considered in a temporal perspective.

16th EMS Annual Meeting 11th European Conference on Applied Climatology (EC AC)

12–16 September 2016 , Trieste, Italy